Active Monitors & DSP Crossovers

[watchnerd]

I'm considering getting a new pair of active monitors.

Almost all my finalist candidates use an analog in (XLR), coupled with a DSP crossover. No digital in.

This means, if I'm listening to digital, it will go through a DA converter, then go through an AD converter, through the DSP, before finally being amplified.

Is this much back and forth DA / AD bad?

And if so, how bad?

And do the benefits of using a DSP-based crossover outweigh the downsides?

 

[Cosmik]

That's interesting. I didn't realise that such monitors didn't provide digital inputs. One reason, I suspect, is that synchronising two remote DSP systems is complicated. With analogue, they can free run and it doesn't matter.

I would bet a large amount of money that no one could hear the difference that an extra AD/DA stage makes (would it even be measurable?). I certainly couldn't hear it. And a DSP crossover brings actual measurable, audible benefits.

 

[Purité Audio]

The more contemporary designs should offer both analogue and digital inputs, look at Grimm or Kii, Genelec even.
Keith

 

[Blumlein 88]

That's interesting. I didn't realise that such monitors didn't provide digital inputs. One reason, I suspect, is that synchronising two remote DSP systems is complicated. With analogue, they can free run and it doesn't matter.

I would bet a large amount of money that no one could hear the difference that an extra AD/DA stage makes (would it even be measurable?). I certainly couldn't hear it. And a DSP crossover brings actual measurable, audible benefits.

If you missed it, you could try an 8th generation vs an original. See if they are audible. The poll has closed, but the files are still up for download. Just don't look past the first post and you won't see the answer.

http://audiosciencereview.com/forum...dac-loop-vs-the-original-can-you-hear-it.448/

 

[watchnerd]

The more contemporary designs should offer both analogue and digital inputs, look at Grimm or Kii, Genelec even.
Keith

Actually, it's not about being contemporary. The Dynaudio LYD series are brand new this year and forego digital input. Eve Audio are also only a few years old and it's the same. It has to do with the fact that most monitor controllers / interfaces have analog out.

Nor is it about necessarily about price point. Barefoot Sound monitors are pretty pricey, the new entry level MM45 being $5595 pair, but only have analog XLR inputs.

 

[watchnerd]

I would bet a large amount of money that no one could hear the difference that an extra AD/DA stage makes (would it even be measurable?). I certainly couldn't hear it. And a DSP crossover brings actual measurable, audible benefits.

Where to measure it would be the other question.

My suspicion is that whatever theoretical differences might be measurable with direct electrical output (e.g. ringing) won't be measurable with a microphone at the speakers because the drivers are so much more distorted / less linear than the upstream signal it would mask everything.

 

[watchnerd]

If you missed it, you could try an 8th generation vs an original. See if they are audible. The poll has closed, but the files are still up for download. Just don't look past the first post and you won't see the answer.

http://audiosciencereview.com/forum...dac-loop-vs-the-original-can-you-hear-it.448/

Interesting, and I'l have a listen, but a bit apples vs pears as there is no DSP.

 

[Fitzcaraldo215]

That's interesting. I didn't realise that such monitors didn't provide digital inputs. One reason, I suspect, is that synchronising two remote DSP systems is complicated. With analogue, they can free run and it doesn't matter.

I would bet a large amount of money that no one could hear the difference that an extra AD/DA stage makes (would it even be measurable?). I certainly couldn't hear it. And a DSP crossover brings actual measurable, audible benefits.

That is actually an interesting point I had not thought of in conjunction with active speakers, each channel having its own separate internal DSP and DACs. Are the master clocks in each separate speaker free enough from random drift from each other? If not, is that potentially audible? Do any manufacturers of actives with digital inputs apply an external I2S synch from a single master clock to all channels to alleviate this?

I have heard, perhaps incorrectly, that much pro digital gear uses syncing to a single, central clock source to alleviate the concerns above. So, there might be some basis for these potential issues.

I suspect that AtoD with analog inputs can be made relatively inaudible. But, I do not like the idea and further suspect that reproduction could be made somewhat more accurate and "transparent" without the extra AtoD. But, obviously, it is simpler and cheaper to just do analog in with AtoD.

 

[RayDunzl]

I'm buying some JBL LSR 308 for portable purposes.

Analog in.

Somebody looked inside and found this:

"While these may be analog input speakers but internally they do all processing & amplification digitally. The crossover & any driver EQ is done in the digital domain. I looked up the major chips that I found on the board.

There is an STA350 capable of up to 50 watts per channel in stereo mode in to a 6 ohm load. This chip handles all the crossover & EQ functions as well. It also can do anti-clipping limiting as well as compression though this last capability is not used here only the anti-clipping limiter is used.

http://www.st.com/web/catalog/sense...568?sc=internet/imag_video/product/251568.jsp

There is a Cirrus Logic 5341 ADC on board as well which handles the analog to digital conversion.

http://www.cirrus.com/en/products/cs5341-42.html

There is also an STM8S0 chip that is a BIOS chip that has a serial interface for storing the parameters used by the STA350. This chip also contains a small 16MHz CPU.

http://www.st.com/web/catalog/mmc/FM141/SC1244/SS1010/LN2/PF251792"

from http://www.head-fi.org/t/779350/review-of-jbl-lsr305

 

[amirm]

Ethan Winer put forward a test a couple of years ago where he had gone through 10+ generations using a cheap PC sound card. I tried to tell the final output and input apart in a first round but failed. I tried sometime later and managed to find it but it was very hard. I think the benefits elsewhere to active drive will far outweigh this difference.

Here is the latest version of that: http://ethanwiner.com/loop-back.htm

 

[watchnerd]

That is actually an interesting point I had not thought of in conjunction with active speakers, each channel having its own separate internal DSP and DACs. Are the master clocks in each separate speaker free enough from random drift from each other? If not, is that potentially audible? Do any manufacturers of actives with digital inputs apply an external I2S synch from a single master clock to all channels to alleviate this?

I have heard, perhaps incorrectly, that much pro digital gear uses syncing to a single, central clock source to alleviate the concerns above. So, there might be some basis for these potential issues.

I suspect that AtoD with analog inputs can be made relatively inaudible. But, I do not like the idea and further suspect that reproduction could be made somewhat more accurate and "transparent" without the extra AtoD. But, obviously, it is simpler and cheaper to just do analog in with AtoD.

I volunteer as a recording engineer assistant for my local symphony and jazz venue (not my profession, but learn a lot).

I've seen master clocks used during recording to keep the multiple AD devices synchronized.

I've never seen master clocks used during mixing or mastering, as there isn't really a point: it's all done on a single DAW, in the digital domain, and whatever jitter was inserted at recording time can't be removed.

The AD converters in the monitor speakers don't affect the file being worked on at the DAW.

The more common complaint is latency.

 

[dallasjustice]

I think the better question for speakers with only analog inputs converted to digital and back is why is that better than a digital input? I can't think of a reason why it would be better. There's no question that the unnecessary AtoD---DSP---DtoA stage will create unnecessary propagation delay. Depending on the design, it could also create an unfixable synchronization problem with the other speakers unless the speakers also have a word clock I/O. There's no good reason for it, IMO.

 

[FrantzM]

I think the better question for speakers with only analog inputs converted to digital and back is why is that better than a digital input? I can't think of a reason why it would be better. There's no question that the unnecessary AtoD---DSP---DtoA stage will create unnecessary propagation delay. Depending on the design, it could also create an unfixable synchronization problem with the other speakers unless the speakers also have a word clock I/O. There's no good reason for it, IMO.

I am with Dallas on this... Why not the digital input and be done? Would cost nothing...
On the JBL LSR 308 , has anyone hacked it to accept digital straight?

 

[Blumlein 88]

I'm buying some JBL LSR 308 for portable purposes.

Analog in.

Somebody looked inside and found this:

"While these may be analog input speakers but internally they do all processing & amplification digitally. The crossover & any driver EQ is done in the digital domain. I looked up the major chips that I found on the board.

There is an STA350 capable of up to 50 watts per channel in stereo mode in to a 6 ohm load. This chip handles all the crossover & EQ functions as well. It also can do anti-clipping limiting as well as compression though this last capability is not used here only the anti-clipping limiter is used.

http://www.st.com/web/catalog/sense...568?sc=internet/imag_video/product/251568.jsp

There is a Cirrus Logic 5341 ADC on board as well which handles the analog to digital conversion.

http://www.cirrus.com/en/products/cs5341-42.html

There is also an STM8S0 chip that is a BIOS chip that has a serial interface for storing the parameters used by the STA350. This chip also contains a small 16MHz CPU.

http://www.st.com/web/catalog/mmc/FM141/SC1244/SS1010/LN2/PF251792"

from http://www.head-fi.org/t/779350/review-of-jbl-lsr305

Thanks for this info Ray. I use three 305's in an AV surround system, and also a pair for monitoring of recordings I do. Nice to know more about how they work. To me another example of how good design and making use of DSP works. These things are far better than $$260/pair speakers have any right to be.

 

[Blumlein 88]

I think the better question for speakers with only analog inputs converted to digital and back is why is that better than a digital input? I can't think of a reason why it would be better. There's no question that the unnecessary AtoD---DSP---DtoA stage will create unnecessary propagation delay. Depending on the design, it could also create an unfixable synchronization problem with the other speakers unless the speakers also have a word clock I/O. There's no good reason for it, IMO.

I too couldn't see how doing the AD is better (though I also doubt it causes any audible harm) compared to straight digital in. I wish it were at least an additional input available to use. Then however who uses it this way?

The market was monitor speakers with balanced inputs. Do you add ADAT (not compatible with Toslink though using the same cable and ports) or do you have both? Do you add coax SPDIF? And if you do two or three or five channels of these do they have digital pass thru? What consumer or prosumer gear has multiple digital outs for people buying $130 speakers? I think for all these reasons considering the market, and remember these sell for $130 each, the way they did it was most cost and performance effective for the widest possible market for a speaker that must have small margins of profit.

 

[fas42]

I would bet a large amount of money that no one could hear the difference that an extra AD/DA stage makes (would it even be measurable?). I certainly couldn't hear it. And a DSP crossover brings actual measurable, audible benefits.

I'm getting the impression that the A/D stage in the Kii Three is a weak link - every video where the Kii has clear audible problems appears to be where it is driven by analogue signal.

 

[Purité Audio]

Spot on how I pleaded with Kii to include a properly designed transparent A/D .
Keith

 

[watchnerd]

I am with Dallas on this... Why not the digital input and be done? Would cost nothing...

What do you do with your turntable?

 

[Purité Audio]

My turntable connects to the phono stage in my Illusonic IAP processor, where like the Kiis A/D is completely transparent.
Keith

 

[Vincent Kars]

In case of digital input the question remains what happens next.
Most DSP chips run at 48 kHz, the more expensive ones like the Kii at 96 kHz
My guess is that all these digital inputs will be resampled to the clock rate of the DSP chip.

Is this better or worse than A/D